This project focuses on a neuroengineering approach to improve rehabilitative strategies involving non-invasive brain stimulation (NIBS) technique called transcranial direct current stimulation (tDCS). tDCS is being investigated heavily in recovery from a variety of neuropsychiatric conditions, including depression and stroke. However, tDCS does not show consistent efficacy across subjects, which can be attributed to inter-individual variability resulting from different strength of electric fields (EF) and resultant changes in neural activity. Such variability in recovery can be decreased if tDCS therapy dose is titrated based on tDCS-generated EF and tDCS- modulated neural activity. Non-invasive methods like scalp electroencephalography (EEG) are convenient to use without involving invasive intracranial recording procedures, e.g., electrocorticography (ECoG) and/or stereoencephalography (SEEG). However, the spatiotemporal resolution of EEG is suboptimal when compared with ECoG/SEEG, probably because of volume conduction. Therefore, a transfer function that can achieve ECoG/SEEG-level precision using EEG recordings is desirable. To that end, this project will establish a framework of simultaneous recording of EEG, ECoG and/or SEEG along with tDCS application. Such a framework will yield a transfer function that may be very useful to investigate tDCS dose-individualization based on neural activity using non-invasive methods (e.g., EEG). Such an approach may be more reliable when compared to a simulation model-based approach. Subjects with refractory epilepsy undergoing ECoG/SEEG implantation serve as a natural model to investigate the real-time reactivity of neural system in response to tDCS. The two overlapping areas to be investigated in this project are: 1. Can EEG-based extrapolations of tDCS-generated EF achieve accuracy comparable to ECoG/SEEG- based extrapolations? ECoG/SEEG have superior spatiotemporal resolution compared to EEG, but are invasive and therefore not practical in stroke subjects. Through analyzing simultaneous recording of both EEG and ECoG/SEEG in subjects, new algorithms will be developed to extrapolate tDCS-generated EF using EEG that can match the accuracy of ECoG/SEEG extrapolations. 2. How tightly correlated are scalp EEG and invasive ECoG/SEEG before/during/after tDCS application? The first step towards understanding direct interaction of tDCS with neural activity is simultaneous administration of tDCS and recording neural activity at various depths. Specialized recording setup is required to achieve this and we plan to use clinical setup with some modifications, ensuring patient safety. Specialized software is required to process the data towards accurate source localization of neural activity during tDCS administration, and to compare EEG-based source localization with the ECoG/SEEG-based. In the long term, our multidisciplinary team is confident to deliver a novel, non-invasive neural feedback-based, neuromodulatory approach to tDCS dose-individualization towards efficacious recovery outcomes.